Synaptojanin 2 Functions at an Early Step of Clathrin-Mediated Endocytosis

Current Biology, Vol. 13, 659–663, April 15, 2003, 2003 Elsevier Science Ltd. All rights reserved. DOI 10.1016/S 09 60 - 98 22 ( 03 )0 0 24 1- 0

Synaptojanin 2 Functions at an Early Step of Clathrin-Mediated Endocytosis Nicole Rusk,1 Phuong U. Le,2 Stefania Mariggio,3 Ginette Guay,2 Cristiano Lurisci,3 Ivan R. Nabi,2 Daniela Corda,3 and Marc Symons1,4,5,* 1 Center for Oncology and Cell Biology North Shore-LIJ Research Institute 350 Community Drive Manhasset, New York 11030 2 Department of Pathology and Cell Biology Universite´ de Montre´al Montre´al, Que´bec H3C 3J7 Canada 3 Department of Cell Biology and Oncology Istituto di Ricerche Farmacologiche Mario Negri Consorzio Mario Negri Sud Via Nazionale 66030 S. Maria Imbaro Italy 4 Department of Surgery North Shore-LIJ Research Institute Manhasset, New York 11030 5 Departments of Surgery and Anatomy & Structural Biology Albert Einstein College of Medicine Bronx, New York 10461

Summary Synaptojanin 2 is a ubiquitously expressed polyphosphoinositide phosphatase that displays a high degree of homology in its catalytic domains with synaptojanin 1 [1, 2]. Neurons of synaptojanin 1-deficient mice display an increase in clathrin-coated vesicles and delayed reentry of recycling vesicles into the fusion-competent vesicle pool, but no defects in early steps of endocytosis [3, 4]. Here we show that inhibition of synaptojanin 2 expression via small interfering (si) RNA causes a strong defect in clathrin-mediated receptor internalization in a lung carcinoma cell line. This inhibitory phenotype is rescued by overexpression of wildtype synaptojanin 2, but not of wild-type synaptojanin 1 or mutant synaptojanin 2 that is deficient in 5ⴕ-phosphatase activity. In addition, electron-microscopic analysis shows that synaptojanin 2 depletion causes a decrease in clathrin-coated pits and vesicles. These results suggest a role for synaptojanin 2 in clathrincoated pit formation and imply that lipid hydrolysis is required at an early stage of clathrin-mediated endocytosis. Taken together, our results also indicate that synaptojanin 2 is functionally distinct from synaptojanin 1. Results and Discussion To examine the role of synaptojanin 2 in clathrin-mediated endocytosis, we employed RNA interference [5] to *Correspondence: [email protected]

inhibit synaptojanin 2 expression. The lung carcinoma line A-549 was transiently transfected with either of two different small interfering (si) RNA oligonucleotides, one directed against a sequence in the 5⬘-phosphatase domain (SJ2 siRNA), the other directed against the 3⬘ untranslated region of the gene (SJ2-3⬘ siRNA). As a control, we used oligos directed against either Rac2, a hematopoietic-specific gene that is not expressed in A-549 cells, or luciferase. No differences were observed between these two controls. Three days after transfection, endogenous synaptojanin 2 protein levels were reduced by 70% ⫾ 10% with SJ2 siRNA and 40% ⫾ 10% with SJ2-3⬘ siRNA compared to control siRNA, as determined by densitometric scanning after Western blot analysis (Figure 1A). We determined the transfection efficiency of siRNA oligos in A-549 cells by using oligos directed against lamin A/C [5]. Analysis by immunofluorescence microscopy showed uniform lowering of lamin A/C expression in more than 90% of cells, whereas the remaining ⬍10% showed normal lamin expression (our unpublished data). We examined the functional consequences of synaptojanin 2 depletion by first testing the effect of SJ2 siRNA on the steady-state level of PI(4,5)P2, a substrate of the 5⬘-phosphatase domain of synaptojanin 2. This lipid plays an important role in clathrin-mediated endocytosis [6, 7] and is an essential binding partner of many endocytic proteins, including adaptor protein 2 (AP2), epsin, clathrin adaptor protein AP180, and dynamin [8, 9]. Its spatio-temporal control is thought to be critical for proper endocytic function [6, 7, 10, 11]. HPLC analysis of A-549 cells metabolically labeled with myo-[3 H]-inositol showed a 20% increase (p ⬍ 0.05) in PI(4,5)P2 levels in SJ2 siRNA-treated cells versus controls, with no significant differences in PI(4)P levels (Figure 1B), supporting a role for synaptojanin 2 in controlling the levels of PI(4,5)P2. Very similar changes were observed in neurons obtained from synaptojanin 1-deficient mice [3]. We next determined the effect of synaptojanin 2 depletion on clathrin-mediated endocytosis. SiRNAtreated cells were incubated in the presence of rhodamine-labeled EGF for 10 min at 37⬚C. Immunofluorescence microscopy showed normal EGF uptake in control siRNA-treated cells, whereas 85% of SJ2 siRNA-treated and 45% of SJ2-3⬘ siRNA-treated cells displayed strong inhibition of EGF internalization (Figure 1C, quantified in Figure 2B). Similar levels of inhibition were observed in early endocytosis when cells were fixed 5 min after incubation with EGF (our unpublished data). Receptormediated transferrin internalization was inhibited to a similar extent (our unpublished data). We also quantified the effect of synaptojanin 2 depletion on receptor internalization by using fluorescence-assisted cell sorting (FACS). SJ2 siRNA treatment did not change the number of EGF receptors on the cell surface in unstimulated conditions, as determined by FACS (our unpublished data), but it caused a 75% decrease in EGF uptake (Figure 1D). These results provide direct evidence for an essential role for synaptojanin 2 at an early step of

Current Biology 660

Figure 1. Synaptojanin 2 Depletion Inhibits Receptor Internalization (A) Synaptojanin 2 siRNA inhibits protein expression. A-549 cells were transfected with 60 pmoles of the indicated siRNA oligos and grown for 72 hr. Cell lysates were analyzed by Western blotting with anti-SJ2 antiserum. Anti-dynamin 2 antiserum was used to demonstrate equal loading. (B) Synaptojanin 2 depletion increases PI(4,5)P2 levels. A-549 cells were transfected with the respective siRNAs, labeled, and analyzed as described in the Experimental Procedures. Data shown are the means ⫾ SEM of three experiments performed in duplicate (*p ⬍ 0.05). (C and D) Synaptojanin 2 depletion inhibits EGF receptor internalization. (C) Fluorescence microscopy. A-549 cells were transfected with the respective siRNAs for 72 hr. After serum starvation, cells were incubated with rhodamine-labeled EGF for 10 min, fixed, and processed for fluorescence microscopy. The top panel shows Dapi staining, and the lower panel shows internalized EGF. The scale bar represents 20 ␮m. The results are representative of six independent experiments performed under the same conditions. (D) FACS analysis. A-549 cells were treated as described for (C). After incubation with EGF, cells were detached and taken up in PBS, and FACS analysis was performed. The black areas in the two left panels represent unstained cells, and the gray areas represent cells transfected with either SJ2 siRNA (top) or control siRNA (bottom). Panels are representative of three independent experiments. The graph shows internalized rhodamine-EGF of SJ2 siRNA-treated cells and controls. Shown is the mean ⫹/⫺ SEM of three independent experiments.

clathrin-mediated endocytosis. In contrast, recent data obtained with neurons from synaptojanin 1-deficient mice shows that loss of synaptojanin 1 does not affect early stages of synaptic vesicle formation but rather slows down the reentry of recycling vesicles into the fusion-competent vesicle pool [4]. To confirm the functional difference between the synaptojanin 1 and 2, we tested whether the expression of recombinant synaptojanin 1 or 2 could restore EGF internalization in cells that are depleted of endogenous synaptojanin 2. To achieve this, we transfected A-549 cells with SJ2 siRNA and, after an incubation time of 48 hr to enable efficient depletion of endogenous synaptojanin 2, we transfected either recombinant synaptojanin 2 or the 170 kDa isoform of synaptojanin 1 and incubated the cells for an additional 24 hr. Unlike the brain-specific

145 kDa isoform of synaptojanin 1, the 170 kDa isoform is also expressed in some nonneuronal cells [12]; we could not, however, detect endogenous 170 kDa synaptojanin 1 in A549 cells by Western blot analysis (our unpublished data). Although SJ2 siRNA also inhibits the expression of recombinant synaptojanin 2 to some extent, the level of exogenous synaptojanin 2, even when expressed in SJ2 siRNA-treated cells, was still significantly higher than the levels of endogenous synaptojanin 2 in untransfected cells (our unpublished data). By this protocol, EGF internalization was restored to normal levels in about 80% of the cells transfected with synaptojanin 2, whereas cells transfected with synaptojanin 1 showed levels of EGF internalization that were similar to those observed in cells treated with SJ2 siRNA alone (Figures 2A and 2B). Transfection of synaptojanin 1 in

Role of SJ2 in Clathrin-Mediated Endocytosis 661

Figure 2. Reconstitution Experiments The 5⬘-phosphatase activity of synaptojanin 2 is essential for clathrin-mediated endocytosis. (A) A-549 cells were transfected with either SJ2 siRNA or control siRNA and 48 hr later were transferred to glass coverslips and transfected with HA-tagged wild-type synaptojanin 2 (wtSJ2), wild-type synaptojanin 1 (wtSJ1), or the 5⬘-phosphatase catalytically inactive synaptojanin 2 (PD*SJ2). After a total of 72 hr, cells were serum starved for 6 hr and incubated with rhodamine-EGF for 10 min. After fixation, cells were processed for indirect immunofluorescence with an anti-HA antibody for synaptojanin 2 or antiserum against synaptojanin 1, followed by a FITC-labeled secondary anti-mouse antibody. Arrows indicate the cell expressing either synaptojanin 1 or PD*SJ2. The scale bar represents 20 ␮m. (B) Quantitative analysis of experiments shown in Figures 1C and 2A. For each condition, 200 cells were scored for their ability to internalize EGF. Shown are means ⫹/⫺ SEM. Data shown are representative of 3 independent experiments.

cells treated with control siRNA did not inhibit EGF internalization, verifying that overexpression of synaptojanin 1 does not in itself inhibit clathrin-mediated endocytosis (Figures 2A and 2B). These results therefore further substantiate a role for synaptojanin 2 in early endocytosis, in contrast to the function of synaptojanin 1. The functional difference between the two synaptojanins is somewhat surprising in light of the fact that they share PI(4,5)P2 as a common substrate [2]. To verify that the role of synaptojanin 2 in endocytosis is indeed due to the catalytic activity of its 5⬘-phosphatase domain, we examined whether a synaptojanin 2 mutant that is catalytically inactive could restore normal EGF internalization in cells depleted of endogenous synaptojanin 2. After treatment of A-549 cells with SJ2 siRNA for 48 hr, we retransfected with a 5⬘-phosphatase-deficient mutant and detected no significant restoration of EGF internalization (Figures 2A and 2B), pointing to the essential role of the 5⬘-phosphatase domain of synaptojanin 2, and therefore PI(4,5)P2 hydrolysis, in the regulation of vesicle internalization. Similar results were obtained with the SJ2-3⬘ siRNA (our unpublished data). To determine at which step of clathrin-mediated endocytosis synaptojanin 2 functions, we performed electron microscopy on siRNA-treated cells. SJ2 siRNA treatment caused a reduction of clathrin-coated pits and vesicles at all stages that we examined. Early and late stages of pit formation were impaired by 50%, whereas clathrin-coated vesicles were reduced by 60% in comparison to control siRNA-treated cells (Figure 3). These data are consistent with our observations that synaptojanin 2 depletion causes a strong inhibition in receptor internalization and confirm a role for synaptojanin 2 at an early stage of clathrin-coated-pit formation. These observations, however, are in contrast with the phenotypes reported for synaptojanin 1-deficient mice and C. elegans that is mutant for unc-26, the synaptojanin ortholog [3, 13]. Neurons of synaptojanin 1-deficient mice display an accumulation of clathrin-coated vesicles, indicating a defect in the uncoating process [3]. A similar phenotype was observed in C. elegans unc-26 [13]. Taken together, these data provide strong evidence for distinct roles of synaptojanins 1 and 2 in clathrinmediated endocytosis. The molecular basis for the different roles of synaptojanins 1 and 2 in clathrin-mediated endocytosis remains to be established. Because the synaptojanins are thought to have the same substrate specificity [2], it is likely that their functional differences are mediated by their specific localization. To date, very little is known about the localization of the two synaptojanins. However, because the small GTPase Rac1 binds to synaptojanin 2, but not to synaptojanin 1 [14, 15], it is tempting to speculate that this interaction contributes to the specific localization and function of synaptojanin 2. Our data also indicate that synaptojanin 2-regulated hydrolysis of PI(4,5)P2 is required for the formation of clathrin-coated pits. We previously showed that overexpression of a membrane-targeted 5⬘-phosphatase domain of synaptojanin 2 also causes inhibition of endocytosis [14]. Taken together, these results underscore the importance of a tight regulation of PI(4,5)P2 levels at the membrane. Either a surplus of PI(4,5)P2, caused by SJ2

Current Biology 662

proteins. A possible candidate for the localized action of synaptojanin 2 is epsin. Epsin binds to PI(4,5)P2 via its epsin NH2-terminal homology (ENTH) domain and has been implicated in an early stage of clathrin-mediated endocytosis [18]. Epsin is thought to facilitate coated-pit formation by promoting positive membrane curvature in conjunction with clathrin polymerization [19]. However, clathrin-coated pits and budding vesicles also display negative curvature at the neck of an emerging vesicle, and localized lipid turnover may be required for the generation of these features [20]. In summary, our results indicate an essential function of synaptojanin 2 in clathrin-coated-pit formation. These findings extend the role of regulated PI turnover in endocytosis from vesicle uncoating to clathrin-coated-pit formation. Experimental Procedures

Figure 3. Synaptojanin 2 Depletion Inhibits Clathrin-Coated-Pit and -Vesicle Formation (A–C) Examples of coated pits: a shallow pit (s) is defined as a pit whose depth is smaller than its width; an invaginated pit (i) is a pit whose depth is equal to or greater than its width. (D) Example of a coated vesicle. The scale bars in (B) and (D) represent 100 nm. The pairs (A,B) and (C,D) have the same magnification. (E) Quantification of clathrin-coated-pit and -vesicle formation. (S) shallow pits; (I) invaginated pits; (CP) total of coated pits; and (CV) coated vesicles. A-549 cells were transfected with the respective siRNAs and, 72 hr after transfection, cells were harvested and prepared for electron microscopy, and the number of clathrin-coated pits (shallow and invaginated) and vesicles was quantified from 20 complete cell profiles as described in the Experimental Procedures. The graphs present data from four independent experiments.

depletion, or a deficit of PI(4,5)P2, caused by overexpression of synaptojanin 2, interferes with clathrin-mediated endocytosis. Studies on the invasion of the bacterial pathogen Salmonella show that the bacterial PI(4,5)P2phosphatase SigD, which shares sequence homology with the 5-phosphatase domain of synaptojanins [16], promotes localized PI(4,5)P2 turnover at the base of the forming phagosome and that this PI(4,5)P2 hydrolysis is essential for vesicle fission [17]. A requirement for PI(4,5)P2 hydrolysis during vesicle formation and fission therefore may be an essential feature of different trafficking processes such as clathrin-mediated endocytosis and phagocytosis. The precise mechanism of action of synaptojanin 2 in clathrin-coated-pit formation remains to be determined. PI(4,5)P2 is a cofactor for a large number of endocytic

Synthesis and Transfection of siRNA siRNAs were synthesized by Dharmacon (CO), and the siRNA duplexes were prepared as instructed by the manufacturer to yield a final concentration of 20 ␮M. The siRNA sequences targeting synaptojanin 2 (accession number AF318616; the synaptojanin 2 isoform we use corresponds to synaptojanin 2B2 as described in Nemoto et al. [15]) correspond to the coding regions 1612–1633 for SJ2 siRNA and 4925–4946 for SJ2-3⬘ siRNA. Control siRNA was generated against GL2 luciferase (accession number X65324), coding region 153–173, or the hematopoietically specific Rac2 (accession number NM002872), coding region 556–577. The accession numbers are from GenBank, and the coding regions are numbered relative to the first nucleotide of the start codon. A-549 cells were plated at 70% confluency in 24-well plates in DMEM supplemented with 10% fetal calf serum without antibiotics 24 hr prior to transfection with 60 pmoles siRNA with oligofectamine (Invitrogen, CA), as described by the manufacturer. For reconstitution experiments, A-549 cells were plated and transfected as described above. Fortyeight hours after transfection with siRNA, cells were replated and transfected with plasmids expressing either HA-tagged wild-type synaptojanin 2 or HA-tagged synaptojanin 2 with a catalytically inactive 5⬘-phosphatase domain (R796A, R803A). Both constructs encode a truncated synaptojanin protein that misses the first 59 amino acids that precede the conserved SAC1 domain. Endocytosis Assay and Immunofluoresence siRNA-transfected cells were grown on coverslips for 72 hr, serum starved for 6 hr, and incubated with 6.5 nM rhodamine-labeled EGF or 50 ␮g/ml rhodamine-labeled transferrin (Molecular Probes, OR) for 10 min. Cells were transferred on ice, acid washed twice for 5 min in 50 mM glycine (pH 3) and 100 mM NaCl, and fixed in 4% formaldehyde (Ted Pella, CA). After being mounted in Vectashield (Vector Laboratories, CA), cells were analyzed via an Olympus IX-70 microscope with a Hamamatsu CCD camera and Inovision software. FACS Analysis Transfection and rhodamine-EGF labeling were carried out as described above. After acid washing, cells were incubated with 5⫻ trypsin for 5 min. Immediately after detachment, cells were taken up in PBS and analyzed by FACS (Facs Calibur, Becton Dickinson, CA) with the Cell Quest software. For binding studies, cells were trypsinized and then incubated with rhodamine-EGF at 4⬚C for 2 hr. Measurement of Intracellular PI(4,5)P2 Levels siRNA transfected A-549 cells were grown for 36 hr, then labeled for another 36 hr in medium 199 (Gibco, MD) containing 12.5 ␮Ci/ ml myo-[3H]inositol (New England Nuclear, MA). Lipids were extracted as described in [21], deacylated at 52⬚C for 45 min in monomethylamine mixture (40% aqueous mono-methylamine:water:nbutyl-alcohol:methanol, 4.5:1:1.125:5.875, v/v). After lyophilization, samples were resuspended in 0.5 ml H2O, to which 0.7 ml n-butyl

Role of SJ2 in Clathrin-Mediated Endocytosis 663

alcohol:petroleum ether (b.p. 40⬚C–60⬚C):ethyl formate (20:4:1, v/v) was added. The lower phase was reextracted and lyophilized. Deacylated lipids were separated by anion exchange HPLC on a Partisil 10SAX column (Jones Chromatography, Mid Glamorgan, UK) [21]. Radioactivity in the eluate was monitored with an online radioactivity flow detector (FLO ONE A-525, Packard, The Netherlands). Electron Microscopy siRNA-transfected A-549 cells were grown for 72 hr, rinsed with 0.1 M sodium cacodylate buffer, and fixed with 2% glutaraldehyde for 60 min at 4⬚C. The fixed cells were rinsed in cacodylate buffer, scraped, and collected by centrifugation. The cell pellet was postfixed in 2% osmium tetroxide at 4⬚C, dehydrated, and embedded in LR-White resin (MecaLab, Que´bec). Ultrathin sections (80 nm) were contrasted with uranyl acetate and lead citrate and examined in a Zeiss CEM902 electron microscope. Complete cell profiles, including a nucleus, were identified at low magnification (3,000⫻), and the plasma membrane was then scanned at high magnification (12,000–30,000⫻) for the presence of clathrin-coated pits or vesicles. Antibodies Antiserum against synaptojanin 1 [12] and synaptojanin 2 [14] was previously described; dynamin 2 antiserum was obtained from Santa Cruz (CA), monoclonal AP2 antibody from Affinity Bioreagents (CO), monoclonal tubulin antibody from Upstate Biotechnology (NY), and monoclonal HA antibody from Roche (IN). Polyclonal antisera against phospho-ERK and ERK as well as secondary HRP-conjugated rabbit or mouse antibodies were obtained from Cell Signaling (MA).

10.

11. 12.

13.

14.

15.

16.

17. Acknowledgments We wish to thank P. McPherson for a gift of synaptojanin 1 antiserum, anonymous reviewers for valuable suggestions, and R. Ruggieri and A. Valster for critical reading of the manuscript. This work was supported by National Institutes of Health grant CA87567 to M.S., by a grant from the Canadian Institutes of Health Research to I.R.N., and by grants from the Italian Association for Cancer Research (AIRC, Milano) and Telethon Italy (n. E.841) to D.C. Received: December 16, 2002 Revised: February 10, 2003 Accepted: February 19, 2003 Published: April 15, 2003 References 1. Nemoto, Y., Arribas, M., Haffner, C., and DeCamilli, P. (1997). Synaptojanin 2, a novel synaptojanin isoform with a distinct targeting domain and expression pattern. J. Biol. Chem. 272, 30817–30821. 2. Khvotchev, M., and Sudhof, T.C. (1998). Developmentally regulated alternative splicing in a novel synaptojanin. J. Biol. Chem. 273, 2306–2311. 3. Cremona, O., Di Paolo, G., Wenk, M.R., Luthi, A., Kim, W.T., Takei, K., Daniell, L., Nemoto, Y., Shears, S.B., Flavell, R.A., et al. (1999). Essential role of phosphoinositide metabolism in synaptic vesicle recycling. Cell 99, 179–188. 4. Kim, W.T., Chang, S., Daniell, L., Cremona, O., Di Paolo, G., and De Camilli, P. (2002). Delayed reentry of recycling vesicles into the fusion-competent synaptic vesicle pool in synaptojanin 1 knockout mice. Proc. Natl. Acad. Sci. USA 99, 17143–17148. 5. Elbashir, S.M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K., and Tuschl, T. (2001). Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494–498. 6. Cremona, O., and De Camilli, P. (2001). Phosphoinositides in membrane traffic at the synapse. J. Cell Sci. 114, 1041–1052. 7. Takei, K., and Haucke, V. (2001). Clathrin-mediated endocytosis: membrane factors pull the trigger. Trends Cell Biol. 11, 385–391. 8. Gillooly, D.J., and Stenmark, H. (2001). Cell biology. A lipid oils the endocytosis machine. Science, 291, 993–994. 9. Cullen, P.J., Cozier, G.E., Banting, G., and Mellor, H. (2001).

18.

19.

20. 21.

Modular phosphoinositide-binding domains—their role in signalling and membrane trafficking. Curr. Biol. 11, R882–R893. Kinuta, M., Yamada, H., Abe, T., Watanabe, M., Li, S.A., Kamitani, A., Yasuda, T., Matsukawa, T., Kumon, H., and Takei, K. (2002). Phosphatidylinositol 4,5-bisphosphate stimulates vesicle formation from liposomes by brain cytosol. Proc. Natl. Acad. Sci. USA 99, 2842–2847. Takei, K., and Haucke, V. (2001). Clathrin-mediated endocytosis: membrane factors pull the trigger. Trends Cell Biol. 11, 385–391. Ramjaun, A.R., and McPherson, P.S. (1996). Tissue-specific alternative splicing generates two synaptojanin isoforms with differential membrane binding properties. J. Biol. Chem. 271, 24856–24861. Harris, T.W., Hartwieg, E., Horvitz, H.R., and Jorgensen, E.M. (2000). Mutations in synaptojanin disrupt synaptic vesicle recycling. J. Cell Biol. 150, 589–600. Malecz, N., McCabe, P.C., Spaargaren, C., Qiu, R., Chuang, Y., and Symons, M. (2000). Synaptojanin 2, a novel Rac1 effector that regulates clathrin-mediated endocytosis. Curr. Biol. 10, 1383–1386. Nemoto, Y., Wenk, M.R., Watanabe, M., Daniell, L., Murakami, T., Ringstad, N., Yamada, H., Takei, K., and De Camilli, P. (2001). Identification and characterization of a synaptojanin 2 splice isoform predominantly expressed in nerve terminals. J. Biol. Chem. 276, 41133–41142. Marcus, S.L., Wenk, M.R., Steele-Mortimer, O., and Finlay, B.B. (2001). A synaptojanin-homologous region of Salmonella typhimurium SigD is essential for inositol phosphatase activity and Akt activation. FEBS Lett. 494, 201–207. Terebiznik, M.R., Vieira, O.V., Marcus, S.L., Slade, A., Yip, C.M., Trimble, W.S., Meyer, T., Finlay, B.B., and Grinstein, S. (2002). Elimination of host cell PtdIns(4,5)P(2) by bacterial SigD promotes membrane fission during invasion by Salmonella. Nat. Cell Biol. 4, 766–773. Chen, H., Fre, S., Slepnev, V.I., Capua, M.R., Takei, K., Butler, M.H., Di Fiore, P.P., and De Camilli, P. (1998). Epsin is an EHdomain-binding protein implicated in clathrin-mediated endocytosis. Nature 394, 793–797. Ford, M.G., Mills, I.G., Peter, B.J., Vallis, Y., Praefcke, G.J., Evans, P.R., and McMahon, H.T. (2002). Curvature of clathrincoated pits driven by epsin. Nature 419, 361–366. Hurley, J.H., and Wendland, B. (2002). Endocytosis: driving membranes around the bend. Cell 111, 143–146. Falasca, M., and Corda, D. (1994). Elevated levels and mitogenic activity of lysophosphatidylinositol in k-ras-transformed epithelial cells. Eur. J. Biochem. 221, 383–389.